Probabilistic Modeling of Heteroscedastic Laboratory Experiments Using Gaussian Process Regression
Publication: Journal of Engineering Mechanics
Volume 144, Issue 6
Abstract
This paper proposes an extension to Gaussian process regression (GPR) for data sets composed of only a few replicated specimens and displaying a heteroscedastic behavior. Because there are several factors that are out of the control of experimenters, it is often impossible to reproduce identical specimens for a same experiment. Moreover, observations from laboratory experiments typically display a heteroscedastic interspecimen variability. Because experiments and specimen manufacturing are expensive, it is uncommon to have more than three specimens to build a model for the observed responses. The method proposed in this paper uses GPR to predict each tested specimen using a shared prior structure and models the global heteroscedastic behavior by combining observations using conjugate prior distributions. An application of the method to high-performance fiber-reinforced concrete experiments highlights fiber addition benefits for reducing water permeability caused by macrocracks.
Get full access to this article
View all available purchase options and get full access to this article.
References
Bansal, A. K., and P. Aggarwal. 2007. “Bayes prediction for a heteroscedastic regression superpopulation model using balanced loss function.” Commun. Stat.: Theory Methods 36 (8): 1565–1575.
Blau, G., M. Lasinski, S. Orcun, S.-H. Hsu, J. Caruthers, N. Delgass, and V. Venkatasubramanian. 2008. “High fidelity mathematical model building with experimental data: A Bayesian approach.” Comput. Chem. Eng. 32 (4): 971–989.
Gardoni, P., A. Der Kiureghian, and K. Mosalam. 2002. “Probabilistic capacity models and fragility estimates for reinforced concrete columns based on experimental observations.” J. Eng. Mech. 128 (10): 1024–1038.
Gardoni, P., K. Nemati, and T. Noguchi. 2007. “Bayesian statistical framework to construct probabilistic models for the elastic modulus of concrete.” J. Mater. Civ. Eng. 19 (10): 898–905.
Gelman, A., J. B. Carlin, H. S. Stern, and D. B. Rubin. 2014. Bayesian data analysis, Vol. 2. New York: Chapman and Hall.
Goldberg, P. W., C. K. Williams, and C. M. Bishop. 1997. “Regression with input-dependent noise: A Gaussian process treatment.” Adv. Neural Inf. Process. Syst. 10: 493–499.
Hubert, M., C. Desmettre, and J.-P. Charron. 2015. “Influence of fiber content and reinforcement ratio on the water permeability of reinforced concrete.” Mater. Struct., 48 (9): 1–13.
Jeon, J.-S., A. Shafieezadeh, and R. DesRoches. 2014. “Statistical models for shear strength of RC beam-column joints using machine-learning techniques.” Earthquake Eng. Struct. Dyn. 43 (14): 2075–2095.
Kersting, K., C. Plagemann, P. Pfaff, and W. Burgard. 2007. “Most likely heteroscedastic Gaussian process regression.” In Proc., 24th Int. Conf. on Machine Learning, 393–400. Corvallis, OR: International Machine Learning Society.
Lampinen, J., and A. Vehtari. 2001. “Bayesian approach for neural networks: Review and case studies.” Neural Networks 14 (3): 257–274.
Le, Q. V., A. J. Smola, and S. Canu. 2005. “Heteroscedastic Gaussian process regression.” In Proc., 22nd Int. Conf. on Machine Learning, 489–496. Bonn, Germany: International Machine Learning Society.
Ma, Y., L. Wang, J. Zhang, Y. Xiang, and Y. Liu. 2014. “Bridge remaining strength prediction integrated with Bayesian network and in situ load testing.” J. Bridge Eng. 19 (10): 04014037.
MacKay, D. 1998. “Introduction to Gaussian processes.” NATO ASI Ser. F Comput. Syst. Sci. 168: 133–166.
Murphy, K. P. 2007. “Conjugate Bayesian analysis of the Gaussian distribution.” Def. 1: 16.
Pal, M., and S. Deswal. 2008. “Modeling pile capacity using support vector machines and generalized regression neural network.” J. Geotech. Geoenviron. Eng. 134 (7): 1021–1024.
Rasmussen, C., and H. Nickisch. 2010. “Gaussian processes for machine learning (GPML) toolbox.” J. Mach. Learn. Res. 11 (Nov): 3011–3015.
Rasmussen, C., and C. Williams. 2006. Gaussian processes for machine learning. Cambridge, MA: MIT Press.
Siddique, R., P. Aggarwal, Y. Aggarwal, and S. Gupta. 2008. “Modeling properties of self-compacting concrete: Support vector machines approach.” Comput. Concr. 5 (5): 123–129.
Słoński, M. 2010. “A comparison of model selection methods for compressive strength prediction of high-performance concrete using neural networks.” Comput. Struct. 88 (21): 1248–1253.
Słoński, M. 2011. “Bayesian neural networks and Gaussian processes in identification of concrete properties.” Comput. Assisted Mech. Eng. Sci. 18 (4): 291–302.
Thiyagarajan, K., and S. Kodagoda. 2016. “Analytical model and data-driven approach for concrete moisture prediction.” In Vol. 33 of Proc., Int. Symp. on Automation and Robotics in Construction (ISARC), 1. Vilnius, Lithuania: Dept. of Construction Economics and Property, Vilnius Gediminas Technical Univ.
Titsias, M. K., and M. Lázaro-Gredilla. 2011. “Variational heteroscedastic Gaussian process regression.” In Proc., 28th Int. Conf. on Machine Learning (ICML-11), 841–848. Bellevue, WA: International Machine Learning Society.
Tolvanen, V., P. Jylänki, and A. Vehtari. 2014. “Expectation propagation for nonstationary heteroscedastic Gaussian process regression.” In 2014 IEEE Int. Workshop on Machine Learning for Signal Processing (MLSP), 1–6. Piscataway, NJ: IEEE.
Vanhatalo, J., J. Riihimäki, J. Hartikainen, P. Jylänki, V. Tolvanen, and A. Vehtari. 2012. “Bayesian modeling with Gaussian processes using the GPstuff toolbox.” Preprint arXiv:1206.5754.
Wang, C. 2014. “Gaussian process regression with heteroscedastic residuals and fast MCMC methods.” Ph.D. thesis, Unive. of Toronto.
Wang, C., and R. M. Neal. 2012. “Gaussian process regression with heteroscedastic or non-Gaussian residuals.” Preprint arXiv:1212.6246.
Yeh, I. C. 2014. “Estimating distribution of concrete strength using quantile regression neural networks.” Appl. Mech. Mater. 584: 1017–1025.
Zhong, J., P. Gardoni, D. Rosowsky, and T. Haukaas. 2008. “Probabilistic seismic demand models and fragility estimates for reinforced concrete bridges with two-column bents.” J. Eng. Mech. 134 (6): 495–504.
Information & Authors
Information
Published In
Journal of Engineering Mechanics
Volume 144 • Issue 6 • June 2018
Copyright
©2018 American Society of Civil Engineers.
History
Received: Jun 22, 2017
Accepted: Dec 14, 2017
Published online: Apr 11, 2018
Published in print: Jun 1, 2018
Discussion open until: Sep 11, 2018
Authors
Metrics & Citations
Metrics
Citations
Download citation
If you have the appropriate software installed, you can download article citation data to the citation manager of your choice. Simply select your manager software from the list below and click Download.